Letter pubs.acs.org/JPCL
Reversible Adsorption of Outer-Sphere Redox Molecules at Pt Electrodes Dileep Mampallil, Klaus Mathwig, Shuo Kang, and Serge G. Lemay* MESA+ Institute for Nanotechnology, University of Twente, Carre 4409, Achterhorst, 7500AE Enschede, The Netherlands S Supporting Information *
ABSTRACT: Adsorption often dominates the response of nanofluidic systems due to their high surface-to-volume ratios. Here we harness this sensitivity to investigate the reversible adsorption of outer-sphere redox species at electrodes, a phenomenon that is easily overlooked in bulk measurements. We find that even though adsorption does not necessarily play a role in the electron-transfer process, such adsorption is nevertheless ubiquitous for the widely used outer-sphere species. We investigate the physical factors driving adsorption and find that this counterintuitive behavior is mediated by the anionic species in the supporting electrolyte, closely following the well-known Hofmeister series. Our results provide foundations both for theoretical studies of the underlying mechanisms and for contriving strategies to control adsorption in micro/nanoscale electrochemical transducers where surface effects are dominant. SECTION: Liquids; Chemical and Dynamical Processes in Solution
A
electrification of the electrode, the presence of other species at the surface and, of course, the identity of the adsorbing species. For example, the degree of specific adsorption of anions on metal electrodes is ion-specific and increases in the order F−, SO42−, Cl−, Br−, I−.4−6 Specific adsorption of ions can in turn change an electrode’s interfacial potential distribution7,8 and potential of zero charge,9,10 with the largest shift arising for the most polarizable ions (e.g., Br− and I−). It has further been argued that specific adsorption of anions can influence electrochemical processes, for example, through the formation of “bridges” between a redox-active species and an electrode that increase the reaction rate.8,11 Given the ubiquity of adsorption, it is reasonable to expect that redox species can also exhibit some level of adsorption at electrodes, even when this is not a necessary step in the electron-transfer process. Here we use the term adsorption in its broadest sense to refer to a total surface excess that includes both physi- and chemisorption. Measuring such adsorption, however, represents a significant challenge because the corresponding surface concentration can correspond to only a small fraction of a monolayer. For example, quartz microbalance experiments were initially interpreted as indicating detectable levels of adsorption, but this interpretation was later disputed when it was suggested that the data could be understood solely by invoking relatively subtle changes in solution density and viscosity caused by redox reactions.12 Less controversially, metal nanoparticles functionalized with ferrocene groups were also found to exhibit adsorption at Pt electrodes.13
n ideal outer-sphere electron-transfer reaction between a molecule in solution and an electrode involves the tunnelling of one or more electrons without significant chemical interactions developing between the two. This is in contrast with inner-sphere reactions that by definition entail some form of linkage between the molecule and the surface − ranging from covalent binding to adsorption at specific sites − before the reaction can proceed.1 Outer-sphere reactions are often associated with fast electron transfer that is relatively independent of the nature of the substrate. Classification into inner- and outer-sphere reactions, however, pertains only to the electron-transfer mechanism. A reaction proceeding along an outer-sphere pathway does not in itself rule out the possibility of residual interactions between the electrode and the molecule(s) involved, insofar as such interactions do not influence the electron-transfer process. A prototypical illustration is the case of a redox species immobilized at an electrode, an approach that was famously employed to quantitatively study electron tunnelling in outersphere reactions.2 One can similarly envision that residual interactions between the reduced or oxidized forms of a redox species and an electrode lead to some degree of weak adsorption. While not directly involved in electron transfer, such adsorption can nonetheless affect electrochemical experiments by influencing mass transport. It is widely accepted that small cations and anions in solution can adsorb onto solid electrodes.3 In general terms, adsorption of a particular species results from a difference in (Gibbs) free energy between the species in bulk solution and at the surface. The microscopic driving mechanism can be enthalpic, entropic, or some combination thereof, and the degree of adsorption is thus sensitive to a broad range of tunable parameters that include temperature, the nature of the solvent, the degree of © 2014 American Chemical Society
Received: November 28, 2013 Accepted: January 27, 2014 Published: January 27, 2014 636
dx.doi.org/10.1021/jz402592n | J. Phys. Chem. Lett. 2014, 5, 636−640
The Journal of Physical Chemistry Letters
Letter
One situation in which even a mild degree of adsorption can have a significant impact on measurements is in micro- or nanofluidic measurement systems. The past decade has seen a surge of activity aimed at creating miniaturized channels and electrodes suitable for electrochemical measurements on minute amounts of solution.14 The motivations for these advances range from fundamental studies of electron transfer and mass transport to the development of miniaturized analytical methods suitable for point-of-care diagnostics.15 An inherent feature of such miniaturized systems is their high surface-to-volume ratio,16 which can greatly magnify the influence of any surface interaction compared with measurements at more conventional macroscopic or micrometer-scale electrodes. Over recent years, our group has performed a series of redox-cycling measurements at pairs of electrodes separated by nanogaps with heights below 100 nm, suggesting that reversible adsorption is ubiquitous in these systems. We find that at high concentrations of redox species adsorption dominates the noise properties of nanoscale sensors18,19 and limits their response time,20 while at ultralow concentrations it significantly decreases the current levels generated by individual molecules.21,22 Quantitatively understanding electrochemical processes in these systems thus requires revisiting the possible influence of adsorption. Here we take advantage of our nanofluidic approach to investigate adsorption of outer-sphere redox couples. The nanogap geometry facilitates the measurements in three distinct manners: (1) it creates a very high surface-to-volume ratio, which increases the relative number of adsorbed molecules and therefore magnifies the influence of adsorption at the surfaces, (2) it ensures that the absolute number of redox molecules in the detection region remains relatively small, which allows stochastic spectroscopy methods to be applied to the signals, and (3) it permits very efficient redox cycling, such that minute amounts of analyte, even down to single molecules at subnanomolar concentrations, can be detected.21,22 These factors render the method extremely sensitive even to small amount of adsorption. Our observations provide a solid basis for understanding the underlying physical mechanisms responsible for the adsorption. A sketch of a nanogap device is shown in Figure 1a. A detailed method for making the devices using optical lithography is given in the Supporting Information (SI), and an optical image of a completed device is shown in Figure 1b. All data shown here employ Pt electrodes based on thin films created by electron-beam evaporation. The active volume, namely, the volume sandwiched between the two redox-cycling electrodes, is defined by the length of the top electrode, La, the width of the bottom electrode, wb, and the separation between the electrodes, h. Here we employ two distinct families of devices, types 1 and 2, with La × wb × h = 100 μm × 3 μm × 130 nm and 10 μm × 3 μm × 60 nm, respectively. The redoxcycling current in a device is given by ilim = (enD/h2)Nsol, where −e is the charge of the electron, n is the number of electrons transferred per cycle, D is the diffusion coefficient, and Nsol is the number of molecules in solution inside the active volume at a given instant. On average, ⟨Nsol⟩ = NALawbhc, where NA is Avogadro’s number and c is the molar concentration of the redox species in the reservoir of the device, which is independent of the level of adsorption at the surface of the electrodes. We first illustrate how reversible adsorption can be directly observed in nanofluidic systems through two separate experi-
Figure 1. (a) Schematic diagram of the longitudinal cross-section of a nanogap device used for redox cycling. The two electrodes are shown in pink. Also identified are the height of the nanochannel, h, the length of the active redox-cycling region, La, and the length of the access channels, Le. (b) Top-view optical image of a device and schematic illustration of its cross-section. The active region (red rectangle) is defined as the volume between the top and the bottom electrodes. The widths of the bottom electrode and of the nanochannel are wb and wt, respectively.
ments, potential-step amperometry and electrochemical correlation spectroscopy, as summarized in Figure 2. Each of these experiments furnishes two independent signatures of adsorption, emphasizing its broad impact on redox-cycling measurements. Potential-step amperometry gives the clearest qualitative signature of adsorption in nanogap devices by exploiting the fact that the degree of adsorption typically depends on the electrode potential. For example, adsorption of the oxidized form of ferrocene is observed to be enhanced with increasing electrode potential.18,20 Abruptly stepping the potential of an electrode to a higher potential therefore causes a temporary decrease in the analyte concentration inside the nanogap as more molecules become bound to the electrode; this local depletion then recovers over time as more molecules diffuse from a large reservoir outside the device. Correspondingly, the redox cycling current, which is proportional to the local concentration of freely diffusing analyte inside the nanodevice, is suppressed immediately following the potential step. It then gradually recovers over time, the rate of recovery being dictated by diffusion from the external reservoir and into the nanodevice. Because this diffusion is hindered by adsorption, slower recovery indicates increased adsorption. Both the magnitude of the initial current suppression and the duration of the recovery are thus indications of the degree of adsorption.20 This phenomenon is illustrated in Figure 2a for 1 mM Fc(MeOH)2 in 1 M KCl supporting electrolyte for a range of temperatures in device type 1. The bottom electrode (reducing) was kept at 0 V. At room temperature (22 °C), a ∼50% suppression of the limiting current is observed upon stepping the potential of the top electrode (oxidizing) from 0 to 0.4 V, indicating enhanced adsorption at the oxidizing electrode. The current then recovers to its diffusion-limited value after ∼20 s. As the temperature is increased, the amplitude of the departure from the limiting current and duration of the transient both decrease, reaching
